Just a reminder of the state of Hawaii

Subject: General Tech | January 10, 2014 - 12:45 PM |
Tagged: radeon, R9 290X, r9 290, hawaii, catalyst, amd

Confirming the results that Ryan and other sites have seen are the results of [H]ard|OCP's testing of two different retail R9 290X GPUs against a pair of press sample cards.  Much as with Ryan's findings even using the newer Catalyst 13.11 Beta 5 driver, Quiet mode performance varies far more than Uber mode does but even Uber mode displays some differences between models.  However they draw a slightly different conclusion based on their experiences, determining that the variance is not just a matter of press samples versus retail cards but a variance between any and all 290X GPUs.  The complexity of this huge chip is such that the differences in manufacturing process and tolerances are to blame and some cards will simply be better than others.  They also are disappointed by AMD's marketing team, citing that the key is 'With NVIDIA GTX 600 and 700 series the video cards are "running faster than advertised" and with AMD R9 290X the video card is running "slower than advertised."'

View Full Size

"The AMD Radeon R9 290X arrived recently with a high level of performance, and a high level of controversy. There have been reports of performance variance between Radeon R9 290X video cards. We have two purchased retail cards today with stock cooling that we will test and see if performance variances exist."

Here are some more Graphics Card articles from around the web:

Graphics Cards

Source: [H]ard|OCP

January 10, 2014 | 01:41 PM - Posted by David (not verified)

"Oh, AMD you are your own worst enemy sometimes".

The same Jeremy Hellstrom just a few weeks back. Most tech sites reported with neutrality or even optimism over Freesync. Not Hellstrom.

Now "something is rotten".

FYI; Sweclockers have looked into the issue of "golden samples" and could not find any. The news out of HARDOCP is that it is about more than that. Well, fair game, but the whole golden sample thing was specious since other tech sites looked into it and couldn't find it.

I rarely accuse writers of bias, in fact almost never, but it is pretty glaring with Hellstrom. And I own an Nvidia card. This site can do better.

January 10, 2014 | 02:44 PM - Posted by Jeremy Hellstrom

"Something is rotten in the state of Denmark." - Hamlet (1.4), Marcellus to Horatio

It was a deliberately misquoted line from Shakespeare.   While I do utilize headlines specifically to grab attention and often a negative sounding quote works better than a positive one; if you read the actual article I am relatively positive in regards to AMD.  I only made positive comments about Freesync, the negative comment was at AMD's inability to market Freesync to enthusiasts as they could have beat NVIDIA's G-SYNC to market three entire generations ago.

Ryan, The Tech Report, Fudzilla and Legit Reviews all found this performance variation (to name the ones off the top of my head).  What I found interesting at [H] was the indication that this is not a Press Sample versus Retail issue but more of a Golden Sample issue, as you mentioned.  With 6.2 billion transistors on this huge chip there is a lot of chance for quality variability, there is nothing inherently evil or misleading about variability but buyers should be aware that it exists.  Again, if AMD had chosen to market this card in a different manner then this would not be perceived as a negative comment but an indication that you might expect some 290X's to be better than others.  Remember when looking for specific serial numbers on Intel processors because some overclocked better than others was a common practice?

Asking for more from AMD is not an indication of bias, unless you are using the definition of having a strong interest in something in which case you would be correct.  That said, I suppose I should avoid certain phrases which the supplier of my last 3 GPUs and some readers might decide they need to take personal offense over.

January 10, 2014 | 03:13 PM - Posted by JohnGR (not verified)

It is not "Press Sample versus Retail issue but more of a Golden Sample issue".

This is where we are starting playing with words to make it look less bad.

"While I do utilize headlines specifically to grab attention and often a negative sounding quote works better than a positive one;"

Yeah, try it with Nvidia or Intel and see what happens. AMD is an easy target when a journalist tries to get more readers.

But to tell you the truth the articles here in PCPer weren't that bad compared to let's say Tom's. I mean they did exaggerate, but not in a way to have to check the authors name to is if it says "Jen-Hsun Huang". I think there aren't articles on the internet with more excitement that AMD did something fishy than Tom's. There was much more honesty in the articles here than bias and a far more mature way to see in to the matter like in the case of CF and frame rating.

January 10, 2014 | 03:56 PM - Posted by Jeremy Hellstrom

You mean the guy who pulled the heatsink off of a running Athlon to prove AMD's chips could catch fire?  Golden samples are a thing, tweaked Press Cards are also a thing but a thing which carries a somewhat negative connotation.

I haven't had a chance to refer to NVIDIA's fondness for rebranding cards to sell to people who don't realize they are just buying last years card at this years prices (or the effect of renaming certain .exe files)  or poke fun at Intel for being monopolizing asses who have lost numerous anti-competition cases ... (or Monster Cables or OnLive or KillerNICs or any of the other manufactures I have ridiculed in the past) but believe you me, when next they do something I disapprove of I will be pointing it out.

I desperately want AMD to survive and since the only thing I can do is offer creative criticism with an attempted humourous twist, that is what I do.  If you want to infer that I hate them and want them to fail there is no way I can stop you but that is not what I am trying to communicate.  Hopefully this year will be a much better one for AMD's flacks.


"I don't know what you mean by 'glory,' " Alice said.
Humpty Dumpty smiled contemptuously. "Of course you don't—till I tell you. I meant 'there's a nice knock-down argument for you!' "
"But 'glory' doesn't mean 'a nice knock-down argument'," Alice objected.
"When I use a word," Humpty Dumpty said, in rather a scornful tone, "it means just what I choose it to mean—neither more nor less."
"The question is," said Alice, "whether you can make words mean so many different things."
"The question is," said Humpty Dumpty, "which is to be master—that's all."

January 10, 2014 | 05:42 PM - Posted by JohnGR (not verified)

Nope I don't think you want AMD to fail, didn't meant that.
I am only saying that the press is easily harsh towards AMD most times and I saw it again with the reviews of the reference cards. Many articles where trying to show a company that was lying, or at best investigate if AMD was lying, and distract the reader's attention from the fact that even with a bad cooling solution, the card was still giving excellent performance, the best performance per dollar. If the problem was at the gpu, no one could do anything about it. All criticism in the world would have been justified because there would be no solution to the problem. But being just a bad quality cooler and not a bad chip, I would expect a more "half full" approach, than a "half empty" one.

The press is usually more careful with Nvidia and Intel. And I think the reason is in your last paragraph. No one will put the word "survive" near Intel's or Nvidia's name.

PS Funny titles at Tom. Still trying to remind to the reader that there is/was a problem with 290X.

PS2 If the new Kaveri does have a problem with high frequencies, I wonder how much the press will use this to paint a picture of a worst than Richland product.

January 16, 2014 | 06:11 AM - Posted by biblicabeebli

Please to not stray from the issue at hand. You directly insulted a tech journalist of some repute, one with many years of experience, and a record of being harsh to companies about their business strategies when they degrade the consumer experience. I've read PCper and and listened to their podcast for a few years now, and Jeremy is not biased. You are wrong.

I will however grant you that there are different interpretations of inconsistencies in the 290/X retail cards and that this jeremy has not covered it from thatangle. Fine.
But you didn't say that.
You said this journalist, and by extension the PCper staff, are bad people.
You are wrong.

January 11, 2014 | 10:04 PM - Posted by Anonymous (not verified)

Jeremy being drunk and belligerent AGAIN!!!

January 16, 2014 | 05:52 AM - Posted by biblicabeebli

nope. you, anonymous, are incorrect. hooray jerks on the internet.

January 10, 2014 | 04:14 PM - Posted by Anonymous (not verified)

Yes Intel and Nvidia have more funds to buy friendly reviews from some of their press lap dogs, and more funds for paid blog FUDers, Astro and Turf, pairs of astroterfers, working the main tech blog sites! But both AMD and Nvidia cherry pick, it's just that AMD's marketing was given to much influnce over AMD's engineering. AMD's engineering probably wanted to lower the clock speeds on the GPUs, but the marketing monkeys were able to convince the CEO to push ahead, instead of lowering clock speeds. Had these GPUs been properly clocked, and the drivers properly written, instead of rushed to market, a lower clocked, cooler running target would have resulted in a properly running consumer refrence design SKU being released, and would have shifted the thremal/varience issues over to the overclocker side of the equation where the Custom OEMs have more resources/$$ to deal with the issue. Marketing should never be given the rudder, and allowed to steer the ship, and I hope AMD has a separate chain of command from engineering to the CEO, beacuse if it does not, that is on Rory's watch.

Man the lifeboats Marketing's At the wheel!!!!

January 10, 2014 | 04:24 PM - Posted by arbiter

The real issue for this, since AMD did an Up to clock with no base clock the card will run at least at. Yes picking best cards for review is not uncommon thing, even nvidia does it but with their min clock the card will provide at least a certain level of performance. Now this wouldn't been such a big problem for AMD if they didn't through that junkie cooler on the card. Although since there was a difference found between the retail card bios and reviewer card sample is another problem. A big side effect though of all this since most all reviewers use an Open Air Test Bench, temps of the gpu tend to be lower then what most end users see in an inclosed case so that up to 10% could be 15-20% on end user. AMD could avoided most this if they spend some $ to make a new cooler.

January 10, 2014 | 04:50 PM - Posted by Anonymous (not verified)

Fail 1: No baseline for minimum performance
Fail 2: Underdimensioned cooling

Those two fails are likely to have been a result of clocks being set higher than the cards were originally designed for.

Instead of redesigning the cooler and delaying the cards, they tried to pull a fast one with marketing and paying off shills.

January 10, 2014 | 04:43 PM - Posted by Anonymous (not verified)

Another example of PCPer going down the toilet.

Cherry picking what they want to project.

January 10, 2014 | 05:17 PM - Posted by Anonymous (not verified)

Well to the last commenter you know you always have a choice of what not to read. Your personal insight really has no bearing to anyone else. I myself find the articles here a lot less biased and more based the facts than some other sites.

January 10, 2014 | 05:40 PM - Posted by Anonymous (not verified)

HARDOCP concludes

"It is obvious that there are differences in performance and clock speed between Radeon R9 290X cards, however those performance differences are minor enough that even the most attentive gamer would not notice these while playing. Given that, the next stop for the two GIGABYTE cards is in our daily driver rig so we can enjoy gaming on an Eyefinity setup."

This quick write up makes it seams like it's a "I told you so" when its not.

It's a Straw Man argument PCPer is making

January 10, 2014 | 06:49 PM - Posted by Jeremy Hellstrom

I do not think that phrase means what you think it means, nor at any point do I see an "I told you so" typed out, in fact I am not sure what it is you think I told you but I am pretty sure your interpretation has nothing with any of the words I actually wrote. 

On the plus side, at least you saw the AMD advertising so it is not a total loss for everyone.

January 10, 2014 | 07:43 PM - Posted by Anonymous (not verified)

Please tell us what David Schroth from [H]ardocp meant to say in his own article, which data and conclusion he wrote haven't already said.

January 10, 2014 | 09:31 PM - Posted by praack

AYI - i am confused now-

so is the r290 not a good card? is it actually on the levelof a 5570?

will it up an die mid game?

is this a new AMD GATE or is it a case of new tech and amd needs a tech to explain what is happening and not sales guys?

does Nvidia have a better idea how to handle it- or have they actually just threw their hands up and under load turned the fan up to full.

i am soo confused/////

January 12, 2014 | 06:30 PM - Posted by Jeremy Hellstrom

The card is amazing, apart from the price problems thanks to the miners; the advertising of the card could have been done a lot better and avoided almost all the concerns may have raised.

January 11, 2014 | 02:49 AM - Posted by snook

a year from now:

reminder the state of Hawaii was variable
Subject: General Tech | January 10, 2015 - 12:45 PM | Jeremy Hellstrom
'With NVIDIA GTX 600 and 700 series the video cards are "running faster than advertised" and with AMD R9 290X the video card is running "slower than advertised."'

the above quoted link to the article is just misleading as to the nature of [H]ard|OCP's conclusion. it was different than PCper/ryan's.

this is actually tiresome, TR is fast becoming my go to...

edit: of course i'm going to keep reading and watching, but, damn let's move forward.

January 11, 2014 | 04:16 AM - Posted by Prodeous

Very nice article.

In short, I read articles from various sources to paint a full picture. I never accept a single sites claim and say other sites are shit. You guys @PCPer fill in my overview of the card.

And every site points to the same problem. GPU cooling is the technical issue for these cards not performing as advertised. Even when they claim up to 1GHz, this only occurs in the first 30 seconds of gameplay, then it never gets back. And reviews of non refernced cooled R9's confirm this fact. Even Toms hardware did a nice article where they replaced the stock cooler with Arctic Accelero Xtreme III, showing how the card bounced back to where it should have been. (also convincing me of getting the R9)

So anyone here who is not happy with the articles at PCPer, here is a few hundred other sites that you can read.

I for one am happy with the quality of this site and it, along with Techreport, are my first two sites that I read for any reviews. Then Anandtech, HardOCP, Xbitlabs, Tomshardware, Techpowerup, Hexus.net, Hardwarecanucks, and any links they provide to other sites like LigitReview or others. Then I can finally make a conclusion for my self if a product is good or not.

PCPer, you guys are awesome and don't Fudging change and stay Canadian Jeremy ;) And hell Ryan, Josh, Allen, Scott and the rest of the team. Keep it up and stay awesome :)

January 11, 2014 | 11:32 AM - Posted by snook

I do read other sites and use the "in article"/"related" links for further information. That's no reason to allow misrepresentation and hyperbole to go unchallenged. Stay Floridian snook :]

January 12, 2014 | 06:21 AM - Posted by Anonymous (not verified)

Sure Nvdia must be funding you guys heavily..So much hate, freaking unbelievable.Sad to say PCPER lost me on this article. You keep doing your thing, but don't misinterpret other sites conclusions. Bullshit article.

January 12, 2014 | 06:31 PM - Posted by Jeremy Hellstrom

Kind of easy to tell who uses adblock isn't it?

January 12, 2014 | 07:14 PM - Posted by Anonymous (not verified)

You must be real ignorant Jeremy. Just because AMD adds show up on PCPer site doesn't mean you don't exhibit any kind of bias towards that product or company. All it says is you took the money for advertising.

It certainly doesn't excuse the pathetic attempts PCPer has displayed in this article especially and various others. For cry'n out loud you tried to put your own spin on it when the author of the article David Schroth and editor and chief Brent Justice of [H]ard|OCP provide there data and conclusions.

It's clear as day with the data they provide. PCPer never bothered to inform its readers if the variances amounted to any king of performance when it tested 2 cards back in November 8.

Talk about misleading or misrepresenting the data to your readers. When other sites where doing actual data collecting on the influences of this variance, what did PCPer do ? By means of Jeremy refer to the Nov 8 article and write up this misrepresentation of [H]ard|OCP article.

Bravo to PCPer for showing how professional they are once more.

January 13, 2014 | 02:58 PM - Posted by Jeremy Hellstrom

Since the poster specifically accused PCPer of taking NVIDIA's money to attack AMD I would suggest that the AMD advertising on our site is a valid thing to point out. 

While it would be nice to be able to predict the future, Ryan did not have a retail card to test until December 18th (http://www.pcper.com/reviews/Graphics-Cards/ASUS-Radeon-R9-290X-DirectCU...) so no, there was no indication on Nov 8th that there were differences between cards.  Once the tests could be conducted they were and we reported on them.  In fact it was [H] who were behind other sites in reporting on the difference in 290 X's which could be why everyone is up in arms when I linked to this reminder that not all 290 X's are equal.

There is nothing I can do to determine what words and meanings you have decided I said, all I can do is discuss the actual words which were typed.  Those would be that [H] found that the variance was not a press sample versus retail tweak but that even the two Gigabyte cards they tested showed variance.  At no point did I type that this amounts to a noticable performance delta, that was your own assumption.  The frequency difference exists, it is up to you to read the actual effect on performance in [H]s' review, or to simply skip straight to their conclusions. 

As to how you can accuse me of putting a spin on an unmodified copied and pasted quote; well that is impressive.


January 13, 2014 | 06:14 PM - Posted by Anonymous (not verified)

PCPer professionalism again!!!

The complexity of this huge chip is such that the differences in manufacturing process and tolerances are to blame and some cards will simply be better than others. They also are disappointed by AMD's marketing team, citing that the key is 'With NVIDIA GTX 600 and 700 series the video cards are "running faster than advertised" and with AMD R9 290X the video card is running "slower than advertised."'

Your inference is some are better then others but FAIL to point out the performance results.

This is the unmodified quote.

"The user thinks in their mind "wow, my video card is actually running faster than advertised." That's the key difference right there. With NVIDIA GTX 600 and 700 series the video cards are "running faster than advertised" and with AMD R9 290X the video card is running "slower than advertised." This is not good for video card perception and reputation."

Your linking to an aftermarket ASUS. These test were done with retail reference cards. What do aftermarket cards have to do with this anyhow?

Now that is impressive!!!

January 13, 2014 | 04:49 PM - Posted by snook

Not as impressive as you leaving this out of your copy/paste

"The user thinks in their mind. "~" That's the key difference right there."

you made it seem that [H] concluded what you copy/pasted, Nv faster than AMD. that's simply not the case.

again, misleading, but standard.

January 14, 2014 | 02:01 PM - Posted by Anonymous (not verified)

Wtf is wrong is with this guy? This PCPER author obviously lacks in comprehension + massively Nvdia biased.

Hello...Site admis... wheres the quality control? don't these article require approval before publication? seems they've been on a vacation since of late!

Unmodified quote from hardOcp.

In every GTX 600 and 700 series video card, even the reference designed ones, we always experienced the real-time in-game clock frequency to exceed the Boost Clock. This made the NVIDIA GPU based video cards look really good to gamers .

The user thinks in their mind "wow, my video card is actually running faster than advertised." That's the key difference right there. With NVIDIA GTX 600 and 700 series the video cards are "running faster than advertised" and with AMD R9 290X the video card is running "slower than advertised." This is not good for video card perception and reputation.

The saving grace in all this is that these performance variances do not affect the gameplay experience or the highest playable settings. In no circumstance are the performance variables so far apart that one video card noticeably performs different than another while actually gaming. Image quality and the experience are exactly the same. What AMD has, rather, is a fundamental challenge in marketing these video cards at 1GHz.

Utterly disappointed with the site.

Fyi i run a gtx 660 on my primary PC.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.